Open Access Anim Biosci Vol. 34, No. 7:1105-1115 July 2021 https://doi.org/10.5713/ajas.19.0884 pISSN 2765-0189 eISSN 2765-0235

Single nucleotide polymorphism-based analysis of the genetic structure of Liangshan pig population

Bin Liu1,2,a, Linyuan Shen1,2,a, Zhixian Guo1,2, Mailing Gan1,2, Ying Chen3, Runling Yang4, Lili Niu1,2, Dongmei Jiang1,2, Zhijun Zhong5, Xuewei Li1,2, Shunhua Zhang1,2,*, and Li Zhu1,2,*

* Corresponding Authors: Objective: To conserve and utilize the genetic resources of a traditional Chinese indigenous Shunhua Zhang Tel: +86-13982086205, Fax: +86-291010, pig breed, Liangshan pig, we assessed the genetic diversity, genetic structure, and genetic E-mail: [email protected] distance in this study. Li Zhu Methods: We used 50K single nucleotide polymorphism (SNP) chip for SNP detection of Tel: +86-13982083385, Fax: +86-2291010, E-mail: zhuli7508@ 163.com 139 individuals in the Liangshan Pig Conservation Farm. Results: The genetically closed conserved population consisted of five overlapping gener­ 1 College of Animal Science and Technology, ations, and the total effective content of the population (Ne) was 15. The whole population Sichuan Agricultural University, Chengdu, Sichuan, 611130, China was divided into five boar families and one non-boar family. Among them, the effective 2 Farm Animal Genetic Resources Exploration size of each generation subpopulation continuously decreased. However, the proportion of and Innovation Key Laboratory of Sichuan polymorphic markers (PN) first decreased and then increased. The average genetic distance Province, Sichuan Agricultural University, Chengdu, Sichuan, 611130, China of these 139 Liangshan pigs was 0.2823±0.0259, and the average genetic distance of the 14 3 Sichuan Province General Station of Animal boars was 0.2723±0.0384. Thus, it can be deduced that the genetic distance changed from Husbandry, Chengdu 610066, China generation to generation. In the conserved population, 983 runs of homozygosity (ROH) 4 Agriculture and Rural Bureau of Mabian Yi were detected, and the majority of ROH (80%) were within 100 Mb. The inbreeding Autonomous County, Mabian, 614600, China 5 Sichuan Academy of Animal Sciences, coefficient calculated based on ROH showed an average value of 0.026 for the whole Chengdu 610066, China population. In addition, the inbreeding coefficient of each generation subpopulation initially increased and then decreased. In the pedigree of the whole conserved population, a These authors contributed equally to this work. the error rate of paternal information was more than 11.35% while the maternal infor­ mation was more than 2.13%. ORCID Conclusion: This molecular study of the population genetic structure of Liangshan pig Bin Liu https://orcid.org/0000-0002-6526-7983 showed loss of genetic diversity during the closed cross-generation reproduction process. Linyuan Shen It is necessary to improve the mating plan or introduce new outside blood to ensure long- https://orcid.org/0000-0001-6072-3268 term preservation of Liangshan pig. Zhixian Guo https://orcid.org/0000-0003-1425-8654 Mailing Gan Keywords: Single Nucleotide Polymorphism (SNP) Chip; Liangshan Pig; Inbreeding https://orcid.org/0000-0001-9900-3559 Coefficient; Genetic Distance; Genetic Diversity Ying Chen https://orcid.org/0000-0002-1580-7259 Runling Yang https://orcid.org/0000-0002-7837-0656 Lili Niu https://orcid.org/0000-0002-9783-0945 Dongmei Jiang INTRODUCTION https://orcid.org/0000-0001-9309-308X Zhijun Zhong According to the data (2004) from the Domestic Animal Diversity Information System https://orcid.org/0000-0002-7640-9545 Xuewei Li (DAD-IS) and Food and Agricultural Organization (FAO), China produces one-third of https://orcid.org/0000-0003-4560-5137 the world's pig breeds [1]. Yet, the number of indigenous breeds has declined sharply in Shunhua Zhang https://orcid.org/0000-0003-0569-0146 the past 20 years due to breeding selection for lean meat and high growth rate of foreign Li Zhu pig breeds. Liangshan pig is a small, traditional Chinese breed mainly distributed in the https://orcid.org/0000-0001-7342-0880 mountain areas of Yi Autonomous Prefecture with an altitude of 1,500 to 2,000 m [2]. It is Submitted Nov 18, 2019; Revised Jan 24, 2020; Accepted Apr 14, 2020 well known for cold tolerance, crude feeding tolerance, and meat quality [3]. Due to a devas­ tating outbreak of African swine fever in 2019, the number of Liangshan pigs declined. Therefore, it is important to study the genetic diversity and the changes in genetic structure of the Liangshan pig population to evaluate and protect China’s abundant genetic resources.

Copyright © 2021 by Animal Bioscience This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, www.animbiosci.org and reproduction in any medium, provided the original work is properly cited. 1105 Liu et al (2021) Anim Biosci 34:1105-1115

Due to its low cost in recent years, genome-wide genotyp­ Polymorphic marker ratio (PN) refers to the proportion of ing has become beneficial when studying and researching polymorphic loci in the target population to the total num­ the genetic information of livestock [4]. The genetic variation ber of loci. We first calculated the minimum allele frequency in some Chinese pig breeds has been extensively studied for each locus using PLINK (v1.90) [7] and then calculated using high-density single nucleotide polymorphism (SNP) PN using a self-programmed R script [9]. We calculated PN chips or whole genome sequencing [5]. Although the cost using the formula as follows: involved in genome-wide sequencing has greatly reduced, it is still relatively expensive. As a result, this hinders the M PN  use of genome-wide sequencing for the analysis of large- N scale samples. In this study, we use 50K SNP chip to analyze genetic diversity, genetic relationship, population structure, where M is the number of sites that exhibit polymorphism and inbreeding coefficient of Liangshan pigs in the conserved and N is the total number of sites. population farm. whereExpected M isheterozygosity the number of (He) sites refers that exhibitto the probabilitypolymorphism of and N is the total number of sites. heterozygosity at any one of the individuals in the popula­ MATERIALS AND METHODS tion; observedExpected heterozygosity heterozygosity (Ho) (He) refers refers to tothe the ratio probability of the of heterozygosity at any one of the number of individuals in a population where a locus is het­ individuals in the population; observed heterozygosity (Ho) refers to the ratio of the number Animal care erozygous to the total number of individuals. When the Ho All animal works were conducted according to the guide­ is lessof individuals than the He, in wea population speculate thatwhere the a population locus is heterozygous has ex­ to the total number of individuals. lines on the care and use of experimental animals established perienced selection or inbreeding; if the Ho is more than the by the Ministry of Agriculture of China. The Animal Care He,When the population the Ho is may less havethan introduced the He, we some speculate other varieties.that the population has experienced selection and Ethics Committee of Sichuan Agricultural University We used PLINK (v1.90) to calculate He and Ho [7]. specifically approved this study under Permit No. DKY-S2017 or inbreeding; if the Ho is more than the He, the population may have introduced some other 6906. Calculation of genetic distance and genetic relationshipvarieties. We used PLINK (v1.90) to calculate He and Ho [7]. Animals We used PLINK (v1.90) to calculate idengtical by state (IBS)

Ear tissues from 139 purebred Liangshan pigs were collected distances and R script to build heat maps. IBS refers to the for DNA extraction from the Liangshan Pig Conservation DNACalculation fragment ofidentical genetic by distance descent inand two genetic or more relationship individ­ Farm of Leshan, Sichuan province. After pedigree data query, uals, and these DNA fragments have the same base sequence. all samples were divided into a total of five generation sub­ IBSWe only used considers PLINK the (v1.90) similarity to calculate of genetic idengtical markers orby alleles state (IBS) distances and R script to build populations (Supplementary Table S1). between individuals, regardless of whether they come from theheat same maps. ancestor IBS orrefers not. toTherefore, the DNA there fragment is no idenneedtical for paby­ descent in two or more individuals, Single nucleotide polymorphism genotyping rental genotyping. The genetic distance based on IBS can DNA was extracted from the ear tissues by phenol-chloro­ stilland analyze these theDNA genetic fragments relationship have the of thesame population base sequence. with­ IBS only considers the similarity of form extraction method [6], and the quality of DNA was out information on the pedigree or ancestral samples. We genetic markers or alleles between individuals, regardless of whether they come from the detected by ultraviolet spectrophotometry (NanoDrop, 2000; used G matrix (v2) and R to calculate kinship values and heat Thermo Scientific, ShangHai, China) and gel electrophoresis mapssame [10]. ancestor G matrix or isnot. a genomic Therefore, relationship there is matrixno need con for­ parental genotyping. The genetic (BIO-RAD & DYPC-31BN, Newbio Gi-1, WuHan, China). structed with whole genome markers. Since the pedigree The qualified 139 DNA samples were genotyped using informationdistance basedof a conserved on IBS populationcan still analyze is usually the not ge recorded,netic relationship of the population without “Zhongxin-I” Porcine Breeding Chip (Beijing Compass G matrix is suitable for calculating the genetic relationship. Agritechnology Co., Ltd., Beijing, China), which contains information on the pedigree or ancestral samples. We used G matrix (v2) and R to calculate 51,315 SNPs. Quality control of genotype data was performed Analysis of population structure using PLINK (v1.90) software [7]. Only autosomal loci were Afterkinship quality values control and heatof the maps genotype [10]. Gdata, matrix 36,592 is a SNPgenomic loci relationship matrix constructed with used; SNPs with a minor allele frequency less than 0.05 and were used to analyze the population structure. The popula­ with a call rate less than 90% were eliminated. tionwhole structure genome is mainly markers. clustered Since by the the pedigree neighboring inform methodation of a conserved population is usually (neighbor-joining [NJ]) and based on IBS distance matrix not recorded, G matrix is suitable for calculating the genetic relationship. Analysis of genetic diversity [11]. Based on the analysis, we can roughly infer which Li­ Effective population size (Ne) refers to the size of an ideal angshan pig individuals are close in blood in general, samples population with the same gene frequency variance or the originated from the same family group. We used PLINK same inbreeding coefficient increment (hybridity attenua­ (v1.90)Analysis for population of population structure structure analysis [7]. tion rate) as the actual population, which is estimated based on the level of linkage disequilibrium [8]. We used SNeP (v1.1) InbreedingAfter quality coefficient control ofanalysis the genotype data, 36,592 SNP loci were used to analyze the software to calculate Ne [8]. Widespread in all populations, runs of homozygosity (ROH) population structure. The population structure is mainly clustered by the neighboring method 1106 www.animbiosci.org (neighbor-joining [NJ]) and based on IBS distance matrix [11]. Based on the analysis, we can

roughly infer which Liangshan pig individuals are close in blood in general, samples originated from the same family group. We used PLINK (v1.90) for population structure

analysis [7].

Inbreeding coefficient analysis

Widespread in all populations, runs of homozygosity (ROH) are contiguous segments in an

Liuindividual et al (2021) genomeAnim Biosci due 34:1105-1115 to complete transfer of a homologous haplotype from a parent to a areprogeny. contiguous The segments length in and an individualfrequency genome of ROH due canto com reflect­ thePedigree group history. accuracy A oflong the ROHconserved population of plete transfer of a homologous haplotype from a parent to a Liangshan pigs indicates recent inbreeding, while a short ROH indicates ancient inbreeding. First, the length progeny. The length and frequency of ROH can reflect the Information on a total of 26 sires and 19 dams were available groupof ROH history. in Aeach long sampleROH indicates was calculated recent inbreeding, by PLINK while (v1.90) [7].in the Then pedigree the ratio record of ofthe 139 total Liangshan pigs (Table 1). After a short ROH indicates ancient inbreeding. First, the length genotypic analysis, 4 wrong sire records were corrected. We of lengthROH in of each the sampleROH fragment was calculated to the bytotal PLINK length (v1.90) of autosomal [7]. genomefound thatwas thecalculated information to get on the 12 sires and 3 dams were er­ Then the ratio of the total length of the ROH fragment to roneous; however, their true parents were not detected via thecoefficient total length of ofinbreeding autosomal based genome on ROH was calculated [12] using to the get following genotyping. formula: A total of 97 sires and 117 dams were excluded the coefficient of inbreeding based on ROH [12] using the from the experimental group tested by the chip. Therefore, it following formula: was impossible to judge the accuracy of the information. Al­ together, in the conserved population, the pedigree error rate k k Length(ROH ) of sire information was more than 11.51%, and the error rate FROH L of dam information was more than 2.16%. where k is the number of ROH in the individual and L is the Genetic diversity of the Liangshan pigs length of the autosomal genome of the species (porcine v10.2 The Ne of these 139 purebred pigs of Liangshan Pig Conser­ where k is the number of ROH in the individual and L is the length of the autosomal genome version of the genome, autosomal length is approximately vation Farm was 15.00 and the proportion of PN was 0.8393. 2,450,713 Kb). We observed significant differences in population effective of the species (porcine v10.2 version of the genome, autosomalcontent length (Ne) amongis approximately the different generations. The sample Pedigree accuracy analysis size of the F5 generation subpopulation was 10 and the ef­ 2,450,713 Kb). Pedigree accuracy was analyzed using Mendelian errors based fective population content was 2.31. The sample size of the on paired individuals (individual-parent, individual-mother) F3 generation subpopulation was 41 and the effective popu­ with genotypes in the pedigree [13]. Mendelian error indi­ lation content was 7.89. Lowest proportion of PN (0.8060) catesPedigree which accuracyof an individual's analysis allele is not from any of its was in the F4 generation subpopulation and highest (0.8393) biological parents, and the Mendelian error rate identifies the was in the F5 generation subpopulation. In the conserved proportionPedigree of accuracy markers thatwas have analyzed made a mistakeusing Mendein the calculian ­errors populationbased on andpaired all theseindividuals five generations, the He (0.3478) lation. If a pair of individuals with a Mendel error rate greater was less than the Ho (0.3551), which suggests some other than(individual-parent, 1% is discovered individual-mother)in the pedigree, then with the genotypespedigree is in the pedigreespecies mixed [13]. inMendelian the population error (Table 2). considered wrong and will be found in other homosexual indicates which of an individual's allele is not from any of its biological parents, and the samples that provide genotypes. If a sample with a Mendel Genetic distance of the conserved population of errorMendelian rate below error 1% rateis not identifies found, thethe individual proportion is ofconsid markers­ thatLiangshan have made pigs a mistake in the ered missing. If multiple samples with a Mendel error rate We measured an IBS distance that varied from 0.1261 to belowcalculation. 1% are found,If a pair the of sample individuals with thewith smallest a Mende Mendell error rate greater0.3442 than for the1% conserved is discovered population in of Liangshan pigs and error rate (born earlier to offspring) is selected as the true the generation subpopulations. We also determined that the parentthe pedigree, of the individual. then the pedigree is considered wrong and will overallbe found population in other averagehomosexual genetic distance to be 0.2823± 0.0259. It showed that the average genetic distance among RESULTSsamples that provide genotypes. If a sample with a Mendel errorLiangshan rate below pigs 1% is far is apart,not found, and the variation of the popula­ tion is quite large. The genetic distance among 14 breeding Singlethe individual nucleotide is polymorphismconsidered missing. characteristics If multiple samples with boarsa Mendel ranged error from rate 0.1659 below to 1%0.3136, and the average genetic A total of 51,315 SNP loci were detected from 139 samples. distance was 0.2723±0.0384. The genetic distances of all the PLINK (v1.90) [7] was used to remove loci on sex chromo­ five generation subpopulations ranged between 0.2765 and somes, and SNP genotyping data of 44,739 SNP loci were used 0.2882. The genetic distance of the F5 generation subpopula­ for the subsequent analysis. tion was the lowest (0.2765±0.0253) and the F4 generation subpopulation was the highest (0.2882±0.0279) (Table 3).

Table 1. Rectifying results of Liangshan pig pedigree Result type Number of sires Number of dams No genotype was provided. 97 117 The original pedigree did not match and the real parents were not detected. 12 3 The original pedigree did not match and the real parents were detected. 4 0 Matching with the original pedigree 26 19

www.animbiosci.org 1107 Liu et al (2021) Anim Biosci 34:1105-1115

Table 2. Genetic diversity parameters of Liangshan pigs by genera- Table 3. Identity by state genetic distance of Liangshan pigs tions Standard Generation Max Min Average Generation Ne PN He Ho deviation F1 6.70 0.8339 0.3526 0.3677 F1 0.3442 01372 0.2826 0.0264 F2 7.78 0.8339 0.3477 0.3604 F2 0.3248 0.1565 02797 0.0218 F3 7.89 0.8146 0.3507 0.3583 F3 0.3363 0.1261 0.2822 0.0271 F4 4.88 0.8060 0.3563 0.3697 F4 0.3395 0.1409 0.2882 0.0279 F5 2.31 0.8393 0.3327 0.3595 F5 0.3052 0.1953 0.2765 0.0253 F 15.00 0.8393 0.3478 0.3551 F 0.3442 0.1261 0.2823 0.0259 Boar 0.3136 0.1659 0.2723 0.0384 Ne, effective population content; PN, the proportion of single nucleotide polymorphisms (SNPs) that displayed polymorphism in 44,739 SNPs Max, maximum value; Min, minimum value. selected from the 50K panel; He, expected heterozygosity; Ho, observed heterozygosity. genetic relationship and some have high genetic relationship. Visualization results of IBS distance matrix and G matrix of These findings indicated a potential inbreeding trend among the conserved population and the different generation sub­ these individuals (Figures 1, 2), and therefore, we need to pay populations showed that most of the individuals have moderate special attention to their mating plan.

Figure 1. Identity by state (IBS) distance matrix of the conserved population of Liangshan pigs. Using Plink software and R package to make IBS heat map. Each small square in IBS distance matrix represents the genetic distance value between two pairs from the first sample to the last sample. The larger the value is, the closer it is to red, that is, the larger the genetic distance between two individuals is, that is, the two individuals are not similar, and vice versa.

1108 www.animbiosci.org Liu et al (2021) Anim Biosci 34:1105-1115

Figure 2. G matrix of the conserved population of Liangshan pigs. Using G matrix software and R package to make G matrix heat map. Each small square represents the value of the relationship between two pairs from the first sample to the last sample. The larger the value is, the closer it is to red, that is, the closer the relationship between two individuals is.

Population family genetic structure of Liangshan pigs families. In addition to the five families, 49 sows were geneti­ Due to the importance of boars in the conservation process, cally less related to any of the tested breeding boars (genetic this study combined the G matrix and IBS distance matrix relationship coefficient less than 0.1). Therefore, they were to analyze the phylogenetic tree of all 14 boars in the popu­ classified as another family (Table 4, Figure 4). lation using the NJ method. The existing boars were divided into five different consanguinity families. Samples labeled Inbreeding coefficient of the Liangshan pigs based on with the same color in the evolutionary tree were evaluated as runs of homozygosity the same family, and the same consanguinity family branches A total of 983 ROH were detected in these 139 Liangshan were divided into the second or more inner clustering units pigs. The total length of ROH of each individual ranged be­ in the evolutionary tree (Figure 3). Based on NJ tree and the tween 2.76 Mb and 522.51 Mb, and the average ROH length relationship matrix, combined with the relationship between of the conserved population was 63.24±101.92 Mb. Liang­ sows and boars of different families, all individuals in the shan pig with ROH less than 100 Mb has the largest number conserved population were eventually divided into five large of individuals (86.11% of the protected population; Figure 5)

www.animbiosci.org 1109 Liu et al (2021) Anim Biosci 34:1105-1115

Figure 3. Evolutionary tree of boar samples in the conserved population. Samples labeled with the same color are evaluated as the same family. All boars include 5 families.

The total number of ROH in each Liangshan pig varied be­ inevitable, which also decrease the accuracy and integrity of tween 0 and 33 (an average of 7.07±6.48), and the individuals pedigrees. In indigenous pig conservation farms, the work­ with 0 to 5 ROH was the largest in the conserved population ing conditions are normally very tough. So far no database (Figure 6). Overall, the number of ROH positively correlated has been established and genetic information relies on the with chromosome length. We obtained the inbreeding coeffi­ handwriting in the records. It is impossible to exclude errors cient value of each individual through ROH statistics. In the in genealogical records caused by the negligence of staff genetically closed conserved population of Liangshan pigs, members or existing unrecorded genealogical information. the average inbreeding coefficient of the conserved popula­ According to statistics, the average pedigree error rate in tion was 0.026. However, the average inbreeding coefficient of animal and tree breeding population can reach up to 10% each generation subpopulation changed significantly. Among [14,15]. The genome homozygosity index based on whole the different generation subpopulations, the highest inbreed­ genomic markers can reflect the inbreeding degree of individ­ ing coefficient was 0.035 in the third generation and the lowest uals. It has more application in revealing the actual inbreeding was 0.018 in the fifth generation (Figure 7). level and inbreeding genetic effect of the population [16]. In the absence of pedigree data, the correction of pedigree DISCUSSION inbreeding coefficient with genome inbreeding coefficient is a more accurate tool to measure individual inbreeding Pedigree accuracy analysis degree [17]. Correcting pedigree errors can improve best In the breeding work, data recording errors or defects are linear unbiased prediction-breeding value predictions and improve heritability estimates [16], which can effectively Table 4. Consanguinity family construction in the conserved popula- protect the Liangshan pig population. Through our analy­ tion sis, we discovered an error rate greater than 11.35% in the Family Gender Quantity paternal information and more than 2.13% in maternal in­ Family A Boar 2 formation on Liangshan pigs, thus meaning that the error Sow 17 rate in paternal information was more than that in maternal Family B Boar 4 information. This may be because the maternal information Sow 11 normally comes from daily farm records on reproduction Family C Boar 2 and mating; however, there are many times in the process Sow 21 of artificial insemination. This may further lead to errors in Family D Boar 4 Sow 32 paternal information. Therefore, we need to strengthen the Family E Boar 2 accuracy of pedigree records in the subsequent breeding Sow 8 works on Liangshan pigs. Other Sow 49

1110 www.animbiosci.org Liu et al (2021) Anim Biosci 34:1105-1115

Figure 4. Partial screenshot of the phylogenetic tree. Samples labeled with the same color are evaluated as the same family.

Genetic diversity generation. Now, the whole conserved population consists Using genotype data to assess the Ne of the current popula­ of five overlapping generations. We recorded an Ne of 15 tion is a very important research hotspot in conservation (the whole conservation population), and with increasing genetics [18,19]. High-density SNP panels improve the ac­ generations, the Ne of each generation subpopulation showed curacy in population parameter estimation, including Ne a significant variation. The Ne of the fifth generation was [20]. Our analysis revealed that the conserved population of only 2.31, which may be because only 10 samples were avail­ Liangshan pigs has remained genetically closed and kept re­ able in this generation. Using the same method, Shin et al producing from the formation of the basic group till the fifth [21] estimated the effective population content of a Landrace

Figure 5. Distribution of ROH lengths in the individuals of Liangshan pigs. Taking 50 SNPs as windows, moving 5 SNPs at a time and calculating the length of ROH. The length of ROH in 0 to 100 Mb is the largest in Liangshan pigs. ROH, runs of homozygosity; SNP, single nucleotide polymor- phism.

www.animbiosci.org 1111 Liu et al (2021) Anim Biosci 34:1105-1115

Figure 6. Distribution of runs of homozygosity (ROH) in the individuals of Liangshan pigs.

pig population (1,128 individuals) as 92.27. Due to the small that the SNP polymorphism (PN) kept varying among differ­ size of the population and their unfavorable natural habitat ent gene ratio subpopulations. The SNP polymorphism (PN) (mountain areas), Liangshan pigs has experienced an extinc­ of the whole conserved population was 0.8393, which was tion crisis. By 2010, the conserved population was established, less than that of other Chinese indigenous breeds (Bihu, Chalu, but the Ne of Liangshan pigs still remains low after five gen­ Chun'an, Jinghua, Shengxianhua, Lanxihua, and Jiaxinghei) erations of closed reproduction due to limited independent [22]. The Ho was more than the He for all the generation sub­ blood families of the basic group. This is currently a major populations and for the whole conserved population, which problem that the vast majority of indigenous livestock and indicates that the conserved population contained mixed- poultry breed in China face. Therefore, in the conservation breeds and need to be further purified by eliminating impure of indigenous pig breeds, we need to actively introduce new individuals. In the conserved population of Liangshan pigs, consanguinity, especially new boar consanguinity, and pay the genetic diversity decreased after generations of closed special attention to selection and mating schemes to prevent reproduction. In addition, the heterozygosity of the popula­ the loss of independent consanguinity in the alternation of tion (both He and Ho) was low, which may be due to inevitable generations. Comprehensive genetic diversity analysis showed inbreeding or loss of less productive families. This is also

Figure 7. Average inbreeding coefficient of the conserved population and different generation subpopulations. By calculating the proportion of the total length of runs of homozygosity (ROH) fragment to the total length of autosomal genome, the coefficient of inbreeding based on ROH was obtained.

1112 www.animbiosci.org Liu et al (2021) Anim Biosci 34:1105-1115 common in other small-sized Chinese indigenous livestock largest. This may be due to the different generations and ­se and poultry breeds [14]. lection strategies used by the conserved farm, as well as the different degrees of selection pressure on the traits of these Genetic distance of the conserved population and the lines [26]. Bosse et al [27] found that ROH larger than 5 Mb generation subpopulations could be detected using a 60K SNP chip, and ROH larger In order to protect the population from the influence of ex­ than 5 Mb were as accurate as the whole genome sequencing. otic varieties, most of the conserved livestock populations in In this experiment, we determined 0 Mb to be the shortest China remain genetically closed. Therefore, it is important ROH of Liangshan pigs and 63.24 Mb as the average length, to study genetic distance, genetic relationship, and genetic which proves the accuracy and validity of our results. In structure of the conserved population, especially the multi- addition, our results showed that 86.11% of ROH were shorter generation overlapping population, to maintain sustainable than 100 Mb, which suggests ancient inbreeding. Due to development [23]. IBS showed that the genetic distance of the limited population size of the indigenous pig breeds and the whole conserved population and of the 14 breeding boars the relatively closed operation system, after reproducing was 0.2823 and 0.2723, respectively. After five generations of for successive generations and the aggravation of overlap­ closed reproduction, the genetic distance of the subpopula­ ping generations, the inbreeding coefficient of the population tions slightly changed, and the fifth generation subpopulation kept increasing and the genetic structure of the conserved had the lowest genetic distance (0.2765). These findings ­in population changed dramatically. Our analysis revealed that dicate that the conserved population experienced a selection the inbreeding coefficient of the genetically closed conserved pressure during generation alternation, which changed the population of Liangshan pigs kept increasing, and the in­ genetic structure of the closed conserved population. breeding coefficient of the third generation subpopulation was the highest (3.5%). Due to the introduction of certain Analysis of population genetic structure boars after that, the inbreeding coefficient decreased. Now, In order to analyze the degree of genetic background differ­ the inbreeding coefficient of the whole conserved popula­ ence of the whole population, we used the NJ method and tion remains at 2.6%. clustered the samples based on IBS distance matrix [5]. In the NJ tree, the whole conserved population was divided CONCLUSION into five different families with boars and a family without boars. At present, the number and inbreeding degree of boars By studying the population genetic structure of Liangshan from each family are not balanced in the whole population. pigs at the molecular level, we found that the effective popu­ In three families, there were only two boars. Therefore, in lation content and the genetic diversity of the conserved the subsequent conservation process, we need to pay special population were low during the process of subculture, which attention to the number of pigs in each family during mating led to a change in the genetic structure of the population to plans to avoid the loss of blood lines and maintain a balanced a certain extent. This variation in genetic structure may be family structure. After identifying the effective blood rela­ caused by other breeds that got mixed in the population. To tionship of breeding boars by SNP chip, we can use this as a protect LiangShan pigs, it is necessary to further purify the guide to select for reproduction and to construct frozen se­ population and increase the matching strength or import men banks of different breeds. This will help reduce the cost foreign consanguinity to ensure long-term preservation of of genetic resources and improve the efficiency of conserva­ genetic resources. Further study on the genetic structure of tion. the local protected population in China at the molecular level will effectively help to protect these excellent local varieties. Analysis of inbreeding coefficient based on runs of homozygosity CONFLICT OF INTEREST Chromosome fragments containing homozygous SNP gen­ otypes are used to infer possible haploids inherited by the We certify that there is no conflict of interest with any financial same individual, and ROH (Froh) is used to estimate the in­ organization regarding the material discussed in the manu­ breeding coefficient of the genome [24]. In addition, a long script. ROH indicates a more recent genetic relationship, a short ROH denotes ancient inbreeding, and a greater quantity of ACKNOWLEDGMENTS ROH fragments implies a higher possibility of inbreeding in the family [25]. At present, the number of individuals with This study was supported by the Sichuan Science and Tech­ ROH length within 100 MB is the largest in the whole popu­ nology Support Program (No. 2016NYZ0050; No. SCSZTD lation, and especially the number of ROH within 5 is the -3-008), the National Natural Science Foundation of China

www.animbiosci.org 1113 Liu et al (2021) Anim Biosci 34:1105-1115

(No. 31530073), and the earmarked fund for China Agricul­ 13. Wiggans GR, VanRaden PM, Bacheller LR, et al. Selection ture Research System (No. CARS-36-05B). and management of DNA markers for use in genomic evalu­ ation. J Dairy Sci 2010;93:2287-92. https://doi.org/10.3168/ REFERENCES jds.2009-2773 14. SanCristobal M, Chevalet C, Haley CS, et al. Genetic diversity 1. Fang M, Hu X, Jiang T, et al. The phylogeny of Chinese indi­ within and between European pig breeds using micro­ genous pig breeds inferred from microsatellite markers. Anim satellite markers. Anim Genet 2006;37:189-98. https://doi. Genet 2005;36:7-13. https://doi.org/10.1111/j.1365-2052. org/10.1111/j.1365-2052.2005.01385.x 2004.01234.x 15. Visscher PM, Woolliams JA, Smith D, Williams JL. Estimation 2. Luo J, Lei H, Shen L, et al. Estimation of growth curves and of pedigree errors in the UK dairy population using micro­ suitable slaughter weight of the Liangshan pig. Asian-Australas satellite markers and the impact on selection. J Dairy Sci J Anim Sci 2015;28:1252-8. https://doi.org/10.5713/ajas.15. 2002;85:2368-75. https://doi.org/10.3168/jds.S0022-0302 0010 (02)74317-8 3. Jones GF. Genetic aspects of domestication, common breeds 16. Munoz PR, Resende MFR, Huber DA, et al. Genomic relation­ and their origin. In: Rothschild MF, Ruvinsky A, editors. The ship matrix for correcting pedigree errors in breeding popul­ genetics of the pig. Wallingford, UK: CAB International; 1998. ations: impact on genetic parameters and genomic selection pp. 17-50. accuracy. Crop Sci 2014;54:1115-23. https://doi.org/10.2135/ 4. Ai H, Yang B, Li J, Xie X, Chen H, Ren J. Population history cropsci2012.12.0673 and genomic signatures for high-altitude adaptation in Tibetan 17. Powell JE, Visscher PM, Goddard ME. Reconciling the pigs. BMC Genomics 2014;15:834. https://doi.org/10.1186/ analy­sis of IBD and IBS in complex trait studies. Nat Rev 1471-2164-15-834 Genet 2010;11:800-5. https://doi.org/10.1038/nrg2865 5. Ai H, Huang L, Ren J. Genetic diversity, linkage disequilibrium 18. Waples RS. Making sense of genetic estimates of effective and selection signatures in Chinese and Western pigs revealed population size. Mol Ecol 2016;25:4689-91. https://doi.org/ by genome-wide SNP markers. PLoS One 2013;8:e56001. 10.1111/mec.13814 https://doi.org/10.1371/journal.pone.0056001 19. Waples RK, Larson WA, Waples RS. Estimating contemporary 6. Chomczynski P, Sacchi N. The single-step method of RNA effective population size in non-model species using linkage isolation by acid guanidinium thiocyanate–phenol–chloroform disequilibrium across thousands of loci. Heredity 2016;117: extraction: twenty-something years on. Nat Protoc 2006;1: 233-40. https://doi.org/10.1038/hdy.2016.60 581-5. https://doi.org/10.1038/nprot.2006.83 20. Wang L, Sørensen P, Janss L, Ostersen T, Edwards D. Genome- 7. Purcell S, Neale B, Todd-Brown K, et al. PLINK: a tool set wide and local pattern of linkage disequilibrium and persis­ for whole-genome association and population-based linkage tence of phase for 3 Danish pig breeds. BMC Genet 2013; analyses. Am J Hum Genet 2007;81:559-75. https://doi.org/ 14:115. https://doi.org/10.1186/1471-2156-14-115 10.1086/519795 21. Shin D, Kim SH, Park J, Lee HK, Song KD. Extent of linkage 8. Barbato M, Orozco-terWengel P, Tapio M, Bruford MW. disequilibrium and effective population size of the Landrace SNeP: a tool to estimate trends in recent effective population population in Korea. Asian-Australas J Anim Sci 2018;31: size trajectories using genome-wide SNP data. Front Genet 1078-87. https://doi.org/10.5713/ajas.17.0237 2015;6:109. https://doi.org/10.3389/fgene.2015.00109 22. Chen J, Peng J, Xiao Q, et al. The genetic diversity and popu­ 9. Chan BKC. Data analysis using R programming. In: Bio­ lation structures of indigenous pig breeds in Zhejiang Province statistics for epidemiology and public health using RAdv. revealed by GGRS sequencing. Anim Genet 2018;49:36-42. New York, USA: Springer Publishing Company; 2015. pp. https://doi.org/10.1111/age.12625 81-154. https://doi.org/10.1891/9780826110268.0003 23. Wang X, Cao HH, Geng SM, Li HB. Genetic diversity of 10 10. VanRaden PM. Efficient methods to compute genomic pre­ indigenous pig breeds in China by using microsatellite dictions. J Dairy Sci 2008;91:4414-23. https://doi.org/10.3168/ markers. Asian-Australas J Anim Sci 2004;17:1219-22. https:// jds.2007-0980 doi.org/10.5713/ajas.2004.1219 11. Kumar S, Stecher G, Li M, Knyaz C, Tamura K. MEGA X: 24. Bjelland DW, Weigel KA, Vukasinovic N, Nkrumah JD. Evalu­ molecular evolutionary genetics analysis across computing ation of inbreeding depression in Holstein cattle using whole- platforms. Mol Biol Evol 2018;35:1547-9. https://doi.org/10. genome SNP markers and alternative measures of genomic 1093/molbev/msy096 inbreeding. J Dairy Sci 2013;96:4697-706. https://doi.org/10. 12. Silió L, Rodríguez MC, Fernández A, et al. Measuring inbreed­ 3168/jds.2012-6435 ing and inbreeding depression on pig growth from pedigree 25. Kirin M, McQuillan R, Franklin CS, Campbell H, McKeigue or SNP‐derived metrics. J Anim Breed Genet 2013;130:349- PM, Wilson JF. Genomic runs of homozygosity record popu­ 60. https://doi.org/10.1111/jbg.12031 lation history and consanguinity. PLoS One 2010;5:e13996.

1114 www.animbiosci.org Liu et al (2021) Anim Biosci 34:1105-1115

https://doi.org/10.1371/journal.pone.0013996 27. Bosse M, Megens HJ, Madsen O, et al. Regions of homozy­ 26. Zanella R, Peixoto JO, Cardoso FF, et al. Genetic diversity gosity in the porcine genome: consequence of demography analysis of two commercial breeds of pigs using genomic and the recombination landscape. PLoS Genet 2012;8: and pedigree data. Genet Sel Evol 2016;48:24. https://doi. e1003100. https://doi.org/10.1371/journal.pgen.1003100 org/10.1186/s12711-016-0203-3

www.animbiosci.org 1115